Relational Dynamics in Earnings Management: Interplay of Celebrity CEOs,CEO-CFO Similarity,and Analyst Coverage

Abstract

This study explores the relational dynamics between the CEO and the CFO that shape the CEO’s engagement in earnings management. Specifically, we investigate how CEO-CFO demographic similarity and analyst coverage interact to influence the CFO’s willingness to comply with the celebrity CEO’s earnings management preferences. Our findings show that the use of discretionary accruals is more pronounced when CFOs share greater demographic similarity with celebrity CEOs; however, this effect diminishes in firms with high levels of analyst coverage. By highlighting CFOs’ compliance as a critical boundary condition that enables celebrity CEOs’ influence over a firm’s financial reporting, this study provides a more nuanced explanation of how executive preferences are enacted within organizations, contributing to upper echelons theory and earnings management research.

JEL CLASSIFICATION: M12; G34

Keywords

celebrity CEOs earnings management CEO-CFO demographic similarity analyst coverage impression management

Introduction

Earnings management has been widely documented in the literature as a form of self-serving behavior by corporate managers (Ali & Zhang, 2015; Bergstresser & Philippon, 2006; Burgstahler & Dichev, 1997; Davidson et al., 2004; Lang et al., 2006; Myers et al., 2007). Under current accounting rules, executives have discretion to increase or decrease reported income, and many do so to serve their own interests rather than those of shareholders. Much of this research has focused on how CEOs’ personal characteristics influence the decision to engage in such practices (Li et al., 2022; Malmendier & Tate, 2005, 2009; Putra & Setiawan, 2024). Yet, corporate leaders do not act in isolation. Their preferences—especially those of powerful or high-status CEOs—must be enacted through interactions with other senior executives who hold operational authority. As Malmendier et al. (2023) note, focusing exclusively on CEOs can lead to misattribution, as it overlooks the influence of other top managers responsible for implementing decisions in their respective domains. In this article, we take a more relational view by focusing on how CEO preferences are shaped and executed through the dynamics of the CEO-CFO dyad, particularly in the context of earnings management.

While there has been considerable research regarding the effect of CEOs on corporate financial reporting, recent attention has turned to the role of CFOs (Ge et al., 2011; Geiger & North, 2006; Ham et al., 2017; Jiang et al., 2010; Qiao et al., 2023). Although the CEO is ultimately responsible for earnings management decisions, CFOs are instrumental and wield considerable influence over the extent of such practices, given their control over the creation and dissemination of financial statements. We contend that CEOs drive the decision to engage in earnings management either through explicit directives or by demanding that the CFOs do whatever is necessary to meet financial expectations. In firms that do engage in earnings management, CFOs are often influenced by the CEO to assent to such practices and are ultimately responsible for their implementation. However, some CFOs may harbor misgivings about implementing aggressive accounting practices and may even resist. Thus, a key question here is, which CFOs will be more easily influenced?

In this article, we use the context of celebrity CEOs to explore the core question of the study. Celebrity CEOs are those CEOs who have gained notoriety in the public press and among their peers for their firm’s performance (e.g., Cho et al., 2016; Wade et al., 2006). Prior research suggests that such CEOs are particularly concerned with protecting their elevated status (Cho et al., 2016; Hayward et al., 2004) and are more likely to engage in earnings management, for example, to meet analysts’ earnings forecasts (Malmendier & Tate, 2009) or to smooth earnings in order to signal financial stability. Consistent with this literature, we propose a baseline hypothesis that celebrity CEOs have a strong motive to engage in earnings management as a form of impression management aimed at maintaining their status.

Financial reporting processes and outcomes, however, are CFO’s primary area of expertise and responsibility. As such, CEOs need compliance or cooperation from the CFO to realize their preference for earnings management. To examine the dynamics between the CEO and the CFO in earnings management, we focus on CEO-CFO relational demography and how it affects the relationship between celebrity status and earnings management. Drawing on extensive research on relational demography and its effect on personal evaluation and attraction (e.g., Byrne, 1997; Montoya & Horton, 2004; Tsui & O’reilly, 1989), we posit that CFOs who are demographically more similar to their CEOs are likely to develop stronger interpersonal attraction toward them, which, in turn, increases their willingness to comply with requests to engage in earnings management. In particular, we build on relational identification theory (Sluss & Ashforth, 2007) to suggest that such attraction may evolve into a deeper sense of identification with the CEO-CFO dyadic relationship, thereby motivating CFOs to comply even in ethically charged decisions.

To better understand the dual role of CEOs and CFOs in a firm’s earnings management, we further posit that CFOs’ compliance behavior should be considered not only as a function of CEO-CFO demographic similarity but also in light of CFOs’ personal concerns about the potential costs and penalties associated with earnings management. Drawing on agency theory, research highlights the critical role of financial analysts as an external governance mechanism in detecting managers’ misbehavior (Healy & Palepu, 2001; Wiersema & Zhang, 2011; Yu, 2008). By increasing the perceived risk of detection and reputational consequences, analyst scrutiny may constrain CFOs’ willingness to acquiesce to earnings management. Accordingly, we propose that CEO-CFO similarity and the firm’s level of analyst coverage interact to jointly influence the CFO’s willingness to comply with the CEO’s earnings management requests. Specifically, we argue that the positive effect of CEO-CFO similarity on earnings management among celebrity CEOs will be attenuated in firms with a high level of analyst coverage because the greater scrutiny amplifies CFOs’ concerns about the risks of engaging in earnings manipulation. A three-way interaction among celebrity CEO, CEO-CFO similarity, and analyst coverage is thus expected.

Our findings demonstrate that the preferences of high-status CEOs are enacted through both internal relational alignment and external monitoring. By adopting this relational-contextual framing, this study contributes primarily to upper echelons theory. Recent work has emphasized the need to clarify “where and how” executive attributes translate into organizational outcomes, particularly through cognitive and relational processes (Neely Jr et al., 2020). Responding to these calls, we show how status-driven impression management, CEO-CFO interpersonal attraction, and analyst scrutiny collectively shape CFO compliance with earnings management, thereby advancing an integrative perspective that links upper echelons theory with complementary lenses such as impression management, relational identification, and agency perspectives. These contributions extend not only to the earnings management literature, but also to broader conversations about how organizational actions reflect the interplay between executive status, internal alignment, and external governance.

The remainder of the article is structured as follows. We begin by establishing a baseline hypothesis that explains why celebrity CEOs become motivated to manage earnings. We then develop our theoretical framework around the interplay between CEO celebrity status, CEO-CFO demographic similarity, and analyst coverage. Next, we describe our data, empirical models, and robustness checks. The article concludes with a discussion of the theoretical and practical implications.

Theory and Hypothesis

Celebrity Status, Impression Management, and Earnings Management

“Celebrity” or “Superstar” status is conferred upon corporate leaders of high-performing firms by prominent business media awards and has attracted growing scholarly attention as a distinctive characteristic that shapes CEO behavior (Cho et al., 2016; Hayward et al., 2004; Lee et al., 2020; Li et al., 2022; Malmendier & Tate, 2005, 2009; Pollock et al., 2024; Shi et al., 2017). While celebrity CEOs gain significant benefits such as higher compensation, prestige, and media visibility (Malmendier & Tate, 2009; Wade et al., 2006), this acclaim often comes with elevated expectations. Compared to non-celebrity CEOs, celebrity CEOs face an ongoing imperative to demonstrate that their success is not merely transient but sustainable, which increases their personal accountability for firm outcomes—creating what prior work refers to as the “burden of celebrity” (Graffin et al., 2008; Hayward et al., 2004; Wade et al., 2006). Perceived threats such as inconsistent performance or a decline in media attention can undermine this status (Lee et al., 2020), increasing their sensitivity to external evaluations and prompting strategic efforts to protect their public image.

Impression management theory posits that individuals actively engage in behaviors to shape and maintain favorable perceptions of themselves (Bolino et al., 2008; Bozeman & Kacmar, 1997; Schlenker, 1980). CEOs, in particular, often employ various impression management tactics during periods of organizational decline or reputational threat (Carberry & King, 2012; Chng et al., 2015), sometimes even at the expense of ethical standards (Harris & Bromiley, 2007). For celebrity CEOs, whose status depends heavily on external validation from the public and other stakeholders, this motivation is especially acute, as they face constant pressure to prove their worth through sustained high performance. Recent studies suggest that such CEOs are more inclined to take bold or risky actions to preserve their public image and meet heightened expectations (Shao et al., 2022; Zhou et al., 2023).

One prominent impression management tactic is earnings management, which refers to the intentional manipulation of accounting rules to influence reported earnings (Davidson et al., 2004). Essentially, earnings management is the “process of manipulating the time profile of earnings” (Fudenberg & Tirole, 1995, pp. 75–76) and involves managerial actions that affect the timing or magnitude of reported income. Because financial statements influence external perceptions of managerial competence, and because managers’ evaluation and compensation are often tied to reported performance, CEOs have strong incentives to manage earnings. For instance, prior research shows that executives can benefit from reduced volatility in earnings, which can make earnings management particularly appealing (Davidson et al., 2004; Zhang et al., 2008).

Prior research identifies earnings management as an impression management tactic used to influence stakeholders’ perceptions of firm performance (Ali & Zhang, 2015; Davidson et al., 2004; Godfrey et al., 2003; Merkl-Davies & Brennan, 2007). For celebrity CEOs, who face constant pressure to maintain favorable impressions, earnings management can serve as a strategic response to the burden of celebrity. Specifically, prior work suggests that managers often prefer to smooth earnings over time or precisely meet performance benchmarks. As noted by Fudenberg and Tirole (1995), this preference reflects the idea that reporting bad earnings should be very costly for CEOs: the cost of reporting earnings even slightly below expectation often outweighs the benefit of reporting earnings above expectations (i.e., the torpedo effect, Bouwman, 2014; Skinner & Sloan, 2002). To mitigate such risks, celebrity CEOs may choose to under-report earnings during high-performing periods, effectively building a strategic reserve to buffer against future underperformance.

We believe that earnings smoothing is particularly attractive to celebrity CEOs who, despite their high performance, carry the “burden of celebrity.” Even slightly missing earnings expectations can disproportionately damage their status. To reduce this risk, celebrity CEOs may deliberately under-report earnings during strong financial periods, building a strategic reserve that can be drawn upon during future downturns to maintain the appearance of consistent success. Supporting this logic, prior studies have documented a positive association between celebrity CEO and earnings management. In particular, Malmendier and Tate (2009) find that celebrity CEOs tend to engage in earnings management to precisely meet analysts’ forecasts, and Li et al. (2022) show that celebrity CEOs are more likely to commit financial misconduct. Thus, our first hypothesis, intended to serve as a baseline that validates and extends prior findings, is as follows:

Hypothesis 1 (Baseline Hypothesis): Celebrity CEOs engage in more earnings management than non-celebrity CEOs.

CEO-CFO Demographic Similarities and CFO’s Compliance to Leadership

Drawing on research on relational demography, we argue that a CFO’s willingness to comply with the CEO’s request for earnings management is shaped by interpersonal attraction, which will particularly arise from demographic similarities between the CEO and the CFO. Prior research shows that individuals tend to exhibit more interpersonal attraction toward each other to the extent that they are similar on salient demographic dimensions (e.g., Montoya & Horton, 2004), and that this interpersonal attraction often leads to more voluntary compliance with requests (Baron, 1971).

The effect of relational demography on individual and organizational outcomes has been examined from various theoretical perspectives. According to self-categorization theory, individuals classify themselves and others into social categories based on salient observable characteristics such as age, gender, race, and education (Brickson, 2000; Stangor et al., 1992; Tajfel & Turner, 1986; Tsui et al., 1995). Individuals focus predominantly on similar and positive features of identified in-group members, while emphasizing the distinct and negative characteristics of out-group members. Through this process of comparison, actors become more socially integrated and attracted to peers sharing the same demographic attributes than out-group members possessing different characteristics (Peccei & Lee, 2005; Tsui et al., 1995).

The similarity-attraction perspective also provides evidence for the tendency of an individual to prefer interacting with and becoming attracted to others who share the same characteristics (Byrne, 1971, 1997). According to this perspective, especially when individuals are involved in a dyadic relationship, demographic similarity leads to perceived similarity in attitudes and values, which, in turn, further enhances interpersonal attraction and trust (e.g., Huang & Iun, 2006; McPherson et al., 2001; Ragins, 1997). Such interpersonal attraction promotes a more favorable attitude toward and positive treatment of the other party. Research that has focused on the vertical dyad (i.e., supervisor and subordinate) supports this logic: greater similarity enhances mutual liking, trust, and cooperative behavior (Loi & Ngo, 2009; Shore et al., 2003; Tsui et al., 2002). Moreover, Tsui et al. (2002) find that demographic similarity within the vertical dyad positively affects the subordinate’s willingness to engage in extra-role behavior, which refers to voluntary actions that go above what is expected or prescribed by their role.

This line of theoretical arguments has been extended to research on corporate governance, which suggests that the similarity-attraction process also operates at the upper echelons of organizations. For instance, CEOs are more likely to appoint new directors who have similar demographic characteristics (Westphal & Zajac, 1995; Zhu & Westphal, 2014), and compensation committees apply less strict performance criteria for CEOs who are demographically similar (Young & Buchholtz, 2002), resulting in higher pay (Belliveau et al., 1996). This research consistently suggests that individuals’ perception of demographic similarity results in more positive dyadic relationships and favoring behaviors toward similar others, since such behaviors “protect, enhance, or achieve a positive social identity” (Tajfel, 1982, p. 24). As Flynn (2005) argues, such identity-based attractions can foster mutual trust and facilitate a reciprocal exchange.

Building on this interpersonal connection, attraction in a dyadic relationship can evolve into relational identification, wherein the subordinate begins to define themselves through the relationship with the superior (Sluss & Ashforth, 2007). In such case, the CFO may come to internalize the dyadic relationship as part of their own identity (e.g., “we rise and fall together”), which blurs the boundary between self and other. This shift in psychological orientation alters the CFO’s evaluative framework: rather than evaluating ethically ambiguous requests from a detached, moral standpoint, the CFO may prioritize preserving the relationship and demonstrating role-based loyalty. In this way, relational identification serves as a critical mechanism through which interpersonal attraction translates into greater behavioral compliance.

Recent research provides further support for this process. Zhong and colleagues (2024) provide meta-analytic evidence that subordinates who develop strong relational identification with their supervisors are more likely to engage in unethical pro-supervisor behavior, that is, behaviors intended to protect or benefit the supervisor. Such actions may include lying to protect their supervisor, exaggerating the supervisor’s performance, and withholding negative information to preserve the supervisor’s reputation (Johnson & Umphress, 2019; Leavitt & Sluss, 2015). Similarly, Steffens et al. (2014) demonstrate that shared social identity serves as the basis for followers’ relational attachment to leaders, amplifying their motivation to maintain alignment with the leader, even at the cost of suppressing their own ethical reservations.

Building on the theoretical arguments and supporting empirical evidence discussed above, we propose that CFOs are more likely to develop interpersonal attraction and a sense of shared identity with their CEO when they are demographically similar, leading to greater voluntary compliance with ethically questionable requests from celebrity CEOs. Specifically, we posit that CEO-CFO similarity in age, gender, ethnicity, tenure, educational background, and functional background will moderate the relationship between CEO celebrity status and earnings management in such a way that the positive relationship between CEO celebrity status and earnings management becomes stronger when the CEO and CFO are similar along the demographic dimensions.

Hypothesis 2: CEO-CFO demographic similarity moderates the relationship between celebrity CEOs and earnings management, such that the positive effect of CEO celebrity on earnings management is stronger when the CEO and CFO are demographically similar.

Analyst Coverage and CFO’s Compliance to Leadership

In the context of financial reporting, recent research suggests that CFOs are held more accountable for accounting misdeeds than CEOs (Feng et al., 2011; Hennes et al., 2008; Mian, 2001). For example, while both CEOs and CFOs may experience job turnover following financial earnings restatement (Arthaud-Day et al., 2006), the turnover rate is substantially higher for CFOs (Hennes et al., 2008). In addition to the labor market consequences, CFOs face significant legal and reputational penalties for engaging in accounting manipulation, including fines, disgorgement of illegal gains, and criminal charges (Feng et al., 2011). Given these potential personal costs, CFOs’ willingness to comply with earnings management demands is likely to depend on how strongly they perceive the risks and consequences of doing so. We argue that financial analyst coverage plays a critical role in shaping these perceptions, primarily by increasing CFO’s perceived risk of detection and by amplifying the reputational consequences of financial misconduct.

First, analyst coverage increases CFOs’ perceived detection risk. As agency theory emphasizes, external governance mechanisms such as analyst coverage help reduce agency problems by increasing managerial transparency and accountability. Consistent with this perspective, prior studies on corporate governance have shown that financial analysts function as an effective external monitoring mechanism (e.g., Jensen & Meckling, 1976) and exert substantial influence over managerial decision-making (Brauer & Wiersema, 2018; Schulz & Wiersema, 2018). For instance, analysts help deter managers from engaging in opportunistic behaviors such as value-destroying acquisitions (Chen et al., 2015) or inefficient R&D spending (Gentry & Shen, 2013).

Furthermore, analysts actively participate in a firm’s information production and distribution process (Healy & Palepu, 2001; Wiersema & Zhang, 2011; Yu, 2008), issuing earnings forecasts and stock recommendations that often include red flags or concerns. They also directly communicate with the firms’ management during earnings calls, frequently raising questions concerning financial irregularities such as abnormal changes in earnings. In most cases, CFOs take responsibility for responding to these specific accounting questions. These ongoing interactions enhance analysts’ ability to detect irregularities in financial reporting (Yu, 2008). The presence of more and better-resourced analysts further increases the probability that earnings manipulation will be identified and exposed. Faced with this heightened risk of detection, CFOs are likely to exercise greater caution when facing pressure from celebrity CEOs to engage in earnings management.

Second, analyst coverage amplifies the reputational consequences of financial misreporting, particularly for CFOs. CFOs are widely perceived as the executives most accountable for the accuracy of financial disclosures and are often the main targets of investor and media scrutiny when irregularities are uncovered (Feng et al., 2011; Hennes et al., 2008). When analysts publicly flag concerns about earnings quality, the resulting reputational fallout is often disproportionately borne by CFOs. These reputational risks are particularly salient given the CFO’s primary role in preparing and defending financial statements. Supporting this view, prior research suggests that CFOs who are more attuned to reputational concerns are less inclined to engage in aggressive earnings management (Gounopoulos et al., 2024) and more likely to resist pressure from CEOs to manipulate earnings to avoid damage to their reputations (Florackis & Sainani, 2021). As such, the reputational consequences associated with high analyst scrutiny may serve as a strong deterrent to compliance with ethically questionable demands from celebrity CEOs.

Taken together, we expect that the heightened risk of detection and reputational consequences will constrain CFOs’ willingness to comply with earnings management requests. Specifically, when analyst coverage is high, CFOs become more attuned to the personal risks of financial manipulation. As a result, even when they share high demographic similarity with their celebrity CEOs, their willingness to comply with such requests is reduced. Based on the arguments above, we thus propose a three-way interaction between celebrity CEO, CEO-CFO demographic similarity, and analyst coverage:

Hypothesis 3: Analyst coverage moderates the positive effect of CEO-CFO demographic similarity on earnings management among celebrity CEOs, such that the effect weakens in firms with a high level of analyst coverage.

Figure 1 illustrates the three-way interaction proposed in the study. Specifically, CEO-CFO demographic similarity strengthens the positive effect of CEO celebrity on earnings management, but this similarity effect is weakened under high levels of analyst coverage.

Figure 1.

Conceptual model: three-way interaction effect of CEO-CFO demographic similarity and analyst coverage on the relationship between celebrity CEO and earnings management.

Methods

Data and Sample

The sample for this study consists of a matching-pair sample of the winners of CEO awards and non-winning CEOs from 1997 to 2011. First, we collected the data for celebrity CEOs from the list of a variety of CEO awards from prestigious business publications and organizations: Business Week, Forbes, Chief Executive, Morningstar.com , and Industry Week. Following the procedure of a nearest-neighbor matching process (Abadie & Imbens, 2011; Malmendier & Tate, 2009), we estimated a logistic regression using observable firm and CEO characteristics that have been used to predict celebrity CEO in literature such as firm size, return on assets, book to market ratio, CEO tenure, and CEO age. The results of the logistic regression reveal a pattern consistent with previous studies (e.g., Malmendier & Tate, 2009): CEOs who lead larger firms, exhibit stronger prior performance, have lower book-to-market ratios, are more tenured, and are relatively younger are more likely to win awards. All variables were statistically significant at the 1% level, except for CEO age, which was significant at the 5% level.

To construct the matching sample of non-winning CEOs, we calculated the propensity scores using the estimates from the logistic model for all firms in the Compustat ExecuComp universe. Then we chose the matching non-winning CEO whose propensity score is closest to that of the winning CEO at the time of the award. To assess the quality of the match, we conducted t-tests on firm and CEO characteristics. The results show no significant differences between award winners and their matched counterparts across most dimensions such as firm size, ROA, profitability, book leverage, CEO age, CEO tenure, CEO duality, CEO ownership, CEO gender, suggesting that the matched pairs are generally comparable. This strengthens the validity of our design by ensuring that the primary difference between the two groups is CEO award status. The final sample consists of 177 celebrity CEOs and 177 non-celebrity CEOs (N = 354) in all industries except finance and insurance for the period 1997 to 2011.

Financial data that was used to measure the earnings management was obtained from the Compustat dataset from 1998 through 2012. We augment the data with information from firms’ proxy statements if the Compustat data are incomplete. Data on CEO and CFO demographic characteristics were obtained from diverse sources such as the BoardEX database, Bloomberg’s Executive Profile & Biography, Notable Names Database (NNDB), Dun and Bradstreet Reference Book of Corporate Management, corporate proxy statements, and annual company reports. We then used the I/B/E/S database to collect data on analysts’ earnings forecasts. A 1-year lagging is used to examine the effect of celebrity status on subsequent earnings management.

Measures

Dependent Variable

Following a majority of studies on earnings management (e.g., Bergstresser & Philippon, 2006; Cohen et al., 2008; Dechow et al., 1995; DeFond & Subramanyam, 1998; Jiang et al., 2010; Jones, 1991; Kothari et al., 2005), we estimated managers’ use of discretionary accruals as a proxy for the magnitude of earnings management. Since it allows managers to use a great deal of discretion in deciding how much revenues and expenses to be shifted across accounting periods, the use of discretionary accruals has been a common way to manage earnings. We particularly used the performance-matched discretionary accruals model (Kothari et al., 2005) to empirically detect earnings management. This approach can control for the effect of firm performance on measured discretionary accruals and thus provide reliable estimates of discretionary accruals. First, we calculated total accruals as the difference between income and cash flows from operations and remove components of accruals that are non-discretionary or beyond the control of the CEO by taking the residual from the regression of total accruals. Specifically, total accruals for firm i in the year t are measured as

\begin{array}{l} T o t a l A c c r u a l_{i, t} = (Δ C A_{i, t} - Δ C L_{i, t} - Δ C a s h_{i, t} \\ + Δ S T D_{i, t} - Δ D e p_{i, t}) / Δ A s s e t s_{i, t - 1} \end{array}

where ∆CA i,t, the change in current assets of firm i from year t − 1 to year t; ∆CL i,t, the change in current liabilities; ∆Cash i,t, the change in cash; ∆STD i,t the change in debt in current liabilities; Dep i t, depreciation and amortization expense in year t; ∆Assets i,t-1, total assets in year t − 1.

Then, the discretionary accrual for firm i in year t was measured by taking the residuals from the following regression, which we estimate using a pooled ordinary least squares (OLS) model with two-digit SIC industry and year fixed effects to control for unobserved industry and year influences

\begin{array}{l} T o t a l A c c r u a l_{i, t} = β_{0} + β_{1 i} (1 / A s s e t s_{i, t - 1}) \\ + β_{2 i} (Δ R e v_{i, t} - Δ A R_{i, t}) + β_{3} (P P E_{i, t}) \\ + β_{4 i} (R O A_{i, t - 1}) + ε_{i, t} \end{array}

where ∆Rev i,t is the change in revenues divided by Assets i,t-1, ∆AR i,t is the change in accounts receivable, divided by Assets i,t-1, PPE i,t is the gross value of property, plant and equipment in year t, divided by Assets i,t-1, ROA i,t-1 is the return on assets in year t.

Following prior studies (Bergstresser & Philippon, 2006; Cohen et al., 2008; Jiang et al., 2010), we use the absolute value of discretionary accruals (ɛi,t) as a proxy for the magnitude of earnings management, which can capture both income-increasing and income-decreasing earnings management¹ that managers may use to smooth earnings over time (e.g., Bouwman, 2014; Burgstahler & Dichev, 1997; DeFond & Park, 1997; Gaver et al., 1995). This approach aligns with our theoretical premise that celebrity CEOs have strong incentives to smooth earnings by making both upward and downward adjustments, thereby reducing unexpected volatility and avoiding surprises that could harm their celebrity status.

Independent Variable

The data for celebrity CEOs were obtained from a list of the winners of CEO awards between 1997 and 2011. The CEO award data were hand-collected from Business Week (“Best Manager”), Forbes (“Best Performing CEOs”), Chief Executive (“CEO of the Year”), Morningstar.com (“CEO of the Year”), and Industry Week (“CEO of the Year”) (see Malmendier and Tate (2009) for more details on each award). After merging this sample with available data from the ExecuComp database, a total of 177 award winners were identified for the 1997–2011 period. Celebrity CEO is then measured as a dichotomous variable where it is coded as 1 if the CEO won an award in focal year t, and 0 otherwise. This approach is well accepted in previous studies on CEO celebrity (Cho et al., 2016; Koh, 2011; Malmendier and Tate, 2009; Wade et al., 2006).

Moderating Variables

Following existing research (e.g., Finkelstein et al., 2009; Westphal & Zajac, 1995; Zhu & Westphal, 2014), a CFO’s demographic similarities with the CEO were measured based on seven major background characteristics such as age, gender, ethnicity, tenure, general and elite educational background, and functional background. To develop measures of overall CEO-CFO similarity, we first created seven measures of similarity. Following prior research (Westphal & Zajac, 1995; Zhu & Westphal, 2014), we coded Age similarity as 1 when their ages differ by less than one standard deviation (7 years, calculated from all CEOs and CFOs in our sample), and 0 otherwise. Gender and Ethnicity similarity (i.e., White, African American, Asian, and Hispanic) was a dichotomous variable, coded as 1 if the CEO and CFO are the same along this dimension, respectively, and 0 otherwise (Zhu & Westphal, 2014). Tenure similarity was measured as the number of years in which the CEO and CFO held the respective executive positions together (Young & Buchholtz, 2002).

We included two types of educational background. Following prior research (Westphal & Zajac, 1995; Zhu & Westphal, 2014), we coded general educational background using four categories: PhD/JD, Master’s, Bachelor’s, or lower. We also used 29 high-prestige institutions and a similar coding approach documented in Finkelstein (1992) to classify elite educational background into three categories: (1) nonelite educational background, (2) undergraduate or graduate school from an elite institution, and (3) both undergraduate and graduate education from elite institutions. General and Elite educational background similarity was then coded as 1 if the CEO and CFO shared the same category, and 0 otherwise.

Finally, CEOs were considered similar with CFOs for functional background when CEOs had previous experience in accounting or finance. Following prior research (Westphal & Zajac, 1995), we assessed a CEO’s functional background based on prior job titles (e.g., treasurer, controller, and/or CFO) and whether the CEO held a professional accounting certification (e.g., CPA and/or CFA). Functional background similarity was coded as 1 if the CEO and CFO belonged to the same category, and 0 otherwise.

To assess the effect of overall CEO-CFO demographic similarity, we created standard scores for each of the seven categories and then combined them into a single index of CEO-CFO similarity (Belliveau et al., 1996; Zhu & Westphal, 2014).² While each demographic dimension may not contribute equally to the overall similarity, the seven demographic characteristics have been theoretically and empirically demonstrated as salient bases for social comparison and categorization among top executives (e.g., Zhu & Westphal, 2014). Therefore, the index is intended as a theoretically informed yet pragmatic composite that captures multiple salient aspects of demographic similarity. In our sample, the CEO-CFO similarity variable takes values ranging from −11.39 (less similar) to 7.08 (more similar), with a mean value of zero.

Following prior studies (e.g., Gentry & Shen, 2013; Jo & Harjoto, 2011), we measured Analyst coverage using the NUMEST variable from the I/B/E/S database. This variable represents the number of distinct financial analysts who have submitted individual earnings per share (EPS) forecasts for a given firm in a given fiscal year (ranging from 1 to 51, with a mean of 17.23). Thus, our measure captures the breadth of analyst following based on the number of analysts included in the consensus forecast for each firm-year, rather than the frequency of forecast updates or total volume of reports. We used the natural logarithm of 1 plus the number of forecasts since it is right-skewed (Bushman et al., 2005; Jo & Harjoto, 2011). For the firms in our sample, the average analyst coverage (log-transformed) ranges from 0.69 to 3.95, with a mean of 2.75.

Control Variables

We controlled for potential confounding factors that may affect firms’ earnings management. Following prior research, we controlled for several firm-specific factors: firm size (Watts & Zimmerman, 1986), measured as the natural log of total assets; market to book ratio (Skinner & Sloan, 2002), measured as market value of assets divided by the book value of assets; book leverage (DeFond & Jiambalvo, 1994), measured as interest-bearing debt divided by total assets; and profitability, measured as operating profit divided by total assets.

We also controlled for governance and CEO-level factors. To control CEOs’ equity incentives that may affect earnings management behavior, we included CEO ownership, measured as the proportion of company stock held by the CEO (Bergstresser & Philippon, 2006) and CEO option ownership, measured as the percentage of options exercisable within 60 days, held by the CEO (Bergstresser & Philippon, 2006). We also include duality, a dummy that equals 1 if the CEO is also the chairman of the board, CEO firm tenure, age, and gender (Malmendier & Tate, 2009).

Finally, because CEOs’ capacity to engage in earnings management in the current period could be affected by the extent to which earnings were borrowed or saved in previous periods (Chung et al., 2002; DeFond & Park, 1997), we included prior earnings management (the absolute value of discretionary accruals in the prior year) in the model to control for this effect. Any unmeasured industry and period-specific effects were also controlled for by including industry and year dummies.

Data Analysis

We used the generalized estimating equations (GEE) method developed by Liang and Zeger (1986) to estimate the impact of a celebrity CEO on earnings management. This method is well suited for analyzing nested data structures, particularly when observations within clusters (e.g., repeated measurements of the same CEO) are correlated, as it accounts for non-independence among observations. In our sample, 37 CEOs received awards more than once during the sample period, resulting in a nested data structure with multiple observations per CEO.³ We specified a Gaussian (normal) distribution for the dependent variable, with an identity link function and an exchangeable correlation structure, and used robust variance estimation (White, 1980) to adjust for within-cluster correlation.

Results

Regression Results

Table 1 reports summary statistics and correlations of all variables we used in this study.

Table 1.

Descriptive Statistics and Correlation Matrix between Variables.

		(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)
1.	Earnings management	1.00
2.	Firm size (log)	–0.14	1.00
3.	Market to book ratio	0.08	−0.06	1.00
4.	Book leverage	−0.10	0.25	0.07	1.00
5.	Profitability	0.01	–0.14	0.14	–0.28	1.00
6.	CEO ownership	0.25	−0.09	0.03	–0.12	−0.05	1.00
7.	CEO option ownership	0.02	–0.32	0.05	−0.03	–0.12	0.01	1.00
8.	CEO age	–0.15	0.11	–0.12	0.13	−0.10	0.02	0.08	1.00
9.	CEO gender	0.02	−0.02	0.00	−0.08	−0.06	−0.01	−0.05	–0.14	1.00
10.	Duality	0.02	0.21	0.00	0.21	–0.12	0.04	0.06	0.18	−0.08	1.00
11.	CEO firm tenure	−0.06	0.18	−0.03	0.11	0.08	0.12	−0.06	0.33	–0.14	0.12	1.00
12.	Prior earnings management	0.23	–0.17	0.18	–0.13	−0.01	0.11	0.05	–0.28	0.09	0.02	–0.20	1.00
13.	CEO-CFO similarity	0.13	0.05	−0.05	0.06	−0.01	0.04	–0.11	–0.16	–0.20	−0.01	0.13	0.01	1.00
14.	Analyst coverage	−0.03	0.40	0.15	–0.16	0.18	–0.20	–0.20	–0.21	0.03	−0.02	–0.12	0.11	0.01	1.00
15.	Celebrity CEO	0.16	0.26	0.17	−0.07	0.07	0.09	−0.09	−0.02	0.02	−0.07	−0.03	0.13	−0.01	−0.01	1.00
	Mean	0.05	8.89	6.66	0.21	0.17	0.02	0.00	55.75	0.01	0.71	20.29	0.05	0.00	2.75	0.50
	S.D.	0.05	1.64	15.21	0.16	0.09	0.05	0.01	7.57	0.12	0.45	11.07	0.06	2.85	0.61	0.50
	N = 354

p < .05 for correlations in bold.

Analysis of the variance inflation factor (VIF) indicates that the mean VIF score is 1.83, with values ranging from 1.17 to 2.84 for the main effects and control variables, suggesting no concerns regarding multicollinearity (Neter et al., 1996). In addition, we examined the VIF scores for all interaction and higher-order terms. Prior to mean-centering, the continuous moderator variable (Analyst coverage), the VIFs for the interaction terms ranged from 10.32 to 11.94, indicating potential multicollinearity issues. To address this, we mean-centered analyst coverage and reconstructed the interaction terms accordingly. After centering, the VIF scores for all variables, including the interaction terms, decreased substantially and remained well below the recommended threshold of 5, further confirming that multicollinearity is not a concern in our analyses.

Table 2 presents the results of the GEE analysis for the relationships among celebrity CEO, earnings management, CEO-CFO demographic similarity, and analyst coverage.

Table 2.

Results of GEE Analysis for the Relationship among CEO Celebrity Status, CEO-CFO Similarity, Analyst Coverage, and Earnings Management.

DV	Earnings management (EM) (Performance-matched absolute discretionary accruals)								Income-decreasing EM (Performance-matched negative discretionary accruals)						Income-increasing EM (Performance-matched positive discretionary accruals)
Variables	Model 1		Model 2		Model 3		Model 4		Model 5		Model 6		Model 7		Model 8		Model 9		Model 10
Constant	0.037		0.049		0.048		0.058		0.051		0.050		0.050		–0.007		–0.008		–0.001
	(0.050)		(0.048)		(0.045)		(0.044)		(0.039)		(0.039)		(0.038)		(0.040)		(0.038)		(0.035)
Firm size (log)	–0.002		–0.003		–0.003		–0.004		–0.003		–0.003		–0.003		–0.001		–0.002		–0.002
	(0.002)		(0.002)		(0.002)		(0.002)		(0.002)		(0.002)		(0.002)		(0.002)		(0.002)		(0.002)
Market to book ratio	0.000		0.000		0.000		0.000		0.000		0.000		0.000		0.000		0.000		0.000
	(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)
Book leverage	0.092	*	0.101	**	0.118	**	0.140	**	–0.072		–0.063		–0.066		0.167	**	0.176	**	0.202	**
	(0.039)		(0.038)		(0.038)		(0.042)		(0.044)		(0.044)		(0.044)		(0.026)		(0.028)		(0.028)
Profitability	0.006		–0.007		–0.021		–0.011		0.006		0.005		0.011		–0.010		–0.018		–0.013
	(0.033)		(0.031)		(0.031)		(0.034)		(0.033)		(0.033)		(0.034)		(0.023)		(0.023)		(0.023)
CEO ownership	0.197	**	0.188	**	0.181	*	0.169	*	0.204	**	0.189	**	0.185	*	–0.038		–0.049		–0.048
	(0.074)		(0.071)		(0.072)		(0.073)		(0.074)		(0.073)		(0.075)		(0.038)		(0.039)		(0.036)
CEO option ownership	–0.082		–0.044		–0.132		–0.157		–0.272		–0.330		–0.300		0.097		0.021		–0.043
	(0.359)		(0.368)		(0.353)		(0.367)		(0.206)		(0.211)		(0.224)		(0.329)		(0.320)		(0.353)
CEO age	0.000		0.000		–0.001		–0.001		–0.001	*	–0.001	*	–0.001	*	0.001		0.001	†	0.001	*
	(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)
CEO gender	0.007		0.006		0.004		0.001		–0.003		–0.001		–0.001		0.018		0.020	†	0.019
	(0.017)		(0.015)		(0.017)		(0.017)		(0.009)		(0.010)		(0.010)		(0.011)		(0.012)		(0.013)
Duality	0.004		0.006		0.004		0.003		0.011	*	0.010	*	0.010	*	–0.004		–0.004		–0.005
	(0.006)		(0.006)		(0.006)		(0.006)		(0.005)		(0.005)		(0.005)		(0.005)		(0.005)		(0.005)
CEO firm tenure	0.000		0.000		0.000		0.000		0.000		0.000		0.000		0.000		0.000		0.000
	(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)		(0.000)
Prior earnings management	0.078		0.073		0.074		0.059		0.000		0.038		0.035		0.038		0.038		0.030
	(0.055)		(0.054)		(0.052)		(0.052)		(0.048)		(0.048)		(0.048)		(0.043)		(0.042)		(0.041)
CEO-CFO similarity	0.003	†	0.003	†	–0.001		0.000		0.002	*	0.000		–0.001		0.001		–0.001		0.000
	(0.002)		(0.002)		(0.002)		(0.002)		(0.001)		(0.001)		(0.001)		(0.001)		(0.002)		(0.002)
Analyst coverage	–0.004		–0.006		–0.005		0.001		–0.006		–0.005		–0.006		–0.001		0.000		0.008
	(0.007)		(0.006)		(0.007)		(0.008)		(0.007)		(0.007)		(0.008)		(0.004)		(0.004)		(0.005)
Industry & Year dummies	Yes		Yes		Yes		Yes		Yes		Yes		Yes		Yes		Yes		Yes
Celebrity CEO			0.020	**	0.019	**	0.022	**	0.015	**	0.014	**	0.014	**	0.003		0.003		0.007
			(0.005)		(0.005)		(0.005)		(0.004)		(0.004)		(0.004)		(0.004)		(0.004)		(0.004)
Celebrity CEO × CEO-CFO similarity					0.007	**	0.006	**			0.004	*	0.005	**			0.004	†	0.003
					(0.002)		(0.002)				(0.002)		(0.002)				(0.002)		(0.002)
Celebrity CEO × Analyst coverage							–0.015	†					0.002						–0.019	*
							(0.009)						(0.007)						(0.008)
CEO-CFO similarity × Analyst coverage							0.004						–0.001						0.005	†
							(0.002)						(0.002)						(0.003)
Celebrity CEO × CEO-CFO similarity							–0.006	*					–0.002						–0.004
× Analyst coverage							(0.003)						(0.002)						(0.004)
Wald chi-squared	152.0	**	247.2	**	852.5	**	435.2	**	87.2	**	126.1	**	171.4	**	122.2	**	176.1	**	200.9	**
R ²	0.257		0.283		0.307		0.325		0.306		0.310		0.319		0.172		0.188		0.222
N	354		354		354		354		354		354		354		354		354		354

Robust standard errors are in parentheses; R² is calculated from OLS regression.

†

p < .10; * p < .05; ** p < .01, two-tailed test.

In Hypothesis 1, we argued that celebrity CEOs will engage in more earnings management than non-celebrity CEOs. The coefficient estimate of celebrity CEO is positive and significant (β = 0.020, p = .000) in Model 2 of Table 2, which indicates that there is a significant positive relationship between celebrity CEO and the absolute discretionary accruals, thus supporting Hypothesis 1. The effect size is also significant: when celebrity CEO changes from −1 SD to +1 SD (from non-celebrity CEO to celebrity CEO), the absolute discretional accrual is increased by about 29.6%. This result provides strong support for the theoretical argument put forth in this article that celebrity CEOs are more likely to engage in earnings management as a means to make the firm’s performance appear less variable to protect their social status and reputation.

In Hypothesis 2, we argued that CEO-CFO demographic similarity would positively moderate the relationship between celebrity CEOs and earnings management. The coefficient on the interaction between celebrity CEO and CEO-CFO similarity in Model 3 of Table 2 is positive and significant (β = 0.007, p = .005), thus supporting Hypothesis 2. We also plotted the simple slopes (i.e., marginal effects) in Figure 2. CEO-CFO similarity took values at ±1 SD from the mean, and celebrity CEO took the value of one and zero. As shown, the marginal effect is stronger at higher similarity levels, suggesting that celebrity CEOs’ preference for earnings management is more likely to be realized when they have developed interpersonal attraction with CFOs through demographic similarity.

Figure 2.

Interactive effect of CEO celebrity status and CEO-CFO similarity on earnings management.

In Hypothesis 3, we predicted a three-way interaction among celebrity CEO status, CEO-CFO demographic similarity, and analyst coverage, such that the positive moderating effect of CEO-CFO similarity on the relationship between celebrity CEOs and earnings management would weaken when analyst coverage is high. The coefficient of the three-way interaction term is negative and significant supporting our prediction, as shown in Model 4 of Table 2 (β = −0.006, p = .034). Following Dawson and Richter (2006), we further conducted the slope difference tests to better interpret this higher-order interaction. High and low values for the moderators were set at one standard deviation above and below the mean.

As shown in Figure 3, the marginal effect (i.e., simple slope) of celebrity CEO status on earnings management is less strongly positive in slope (1), which represents high CEO-CFO similarity and high analyst coverage, than in slope (2), which represents high CEO-CFO similarity and low analyst coverage. This slope difference is statistically significant (t = −2.984, p = .003), as reported in Table 3. In addition, the estimated marginal effect shows that the presence of a celebrity CEO increases earnings management by approximately 5.7 percentage points when analyst coverage is low (slope (2)), but only by about 2.0 percentage points when analyst coverage is high (slope (1)), implying a substantive reduction of roughly 3.7 percentage points. These results provide robust support for Hypothesis 3, illustrating how analyst coverage attenuates the positive moderating role of CEO-CFO similarity.

Figure 3.

Three-way interaction effect of CEO celebrity status, CEO-CFO similarity, and analyst coverage on earnings management.

Table 3.

Slope Differences for the Three-Way Interaction.

Pair of slopes	t for slope difference	p for slope difference
(1) and (2)	–2.984	0.003
(1) and (3)	1.033	0.302
(1) and (4)	0.950	0.343
(2) and (3)	2.964	0.003
(2) and (4)	3.402	0.001
(3) and (4)	–0.011	0.991

Additional Analyses and Robustness Checks

We further tested our hypotheses using alternative measures of the dependent and moderating variables, and most of the results remain the same. First, following Hazarika et al. (2012), we separated the earnings management variable as income-increasing and income-decreasing earnings management, and re-estimated all regression models to see the effects of celebrity status on the two opposite directions for earnings management. Income-increasing (decreasing) earnings management was measured as performance-matched positive (negative) discretionary accruals. As shown in Models 5 and 8 of Table 2, respectively, the coefficient of celebrity CEO for income-decreasing earnings management is positive and statistically significant (β = 0.015, p = .001), whereas that for income-increasing earnings management is not statically significant. The result also shows the positive and significant moderating effect of CEO-CFO demographic similarity on income-decreasing earnings management (β = 0.004, p = .039) in Model 6. We discuss the implications of this result in more detail in the discussion section.

Second, we used an alternative measure of CEO-CFO similarity to test the effect of CEO-CFO similarity on the relationship between celebrity CEOs and earnings management. Following Zhu and Westphal (2014), we combined the seven similarity measures into a single index of CEO-CFO similarity by using principal components analysis. In this analysis, the CEO-CFO similarity variable takes values ranging from −4.5 (less similar) to 3.6 (more similar), with a mean value of zero. The results using this re-estimation were unchanged although the significance level dropped slightly (β = 0.007, p = .055).

Third, prior research has provided theoretical and empirical evidence that each of the demographic dimensions such as age, gender, ethnicity, tenure, educational, and functional background is a salient base for determining demographic similarity. Thus, we tested the moderating effects of CEO-CFO similarity using each of the seven attributes instead of the combined index of CEO-CFO similarity. We found significant and positive effects for the similarity in gender (β = 0.041, p = .053), tenure (β = 0.002, p = .084), elite educational background (β = 0.024, p = .043), and functional background (β = 0.037, p = .022). However, the moderating effect of age similarity and general educational background similarity were not supported in this analysis although they were in the same direction as hypothesized. The moderation effect of ethnicity similarity exists in the opposite direction than was expected (β = −0.032, p = .010). In sum, these additional analyses indicate the robustness of our results.

Finally, as an additional falsification check, we conduct a placebo test to verify that CEO celebrity status does not influence non-discretionary accruals, which are mechanically determined by firm fundamentals and should not be subject to managerial discretion. Specifically, following Kothari et al. (2005), we estimate a performance-matched accruals model and obtain the fitted values, which capture non-discretionary accruals. We then regress these fitted values on CEO celebrity status using the same set of control variables as in our main analyses. The results show no significant relationship between CEO celebrity and non-discretionary accruals (Coef. = 0.0005, robust SE = 0.0035, p = .881), providing additional support for the validity and robustness of our findings.

Discussion

This study examines how relational dynamics between the CEO and CFO influence the enactment of CEO preferences in the financial reporting process. Drawing on a relational perspective, we theorize and find that CEO-CFO demographic similarity increases the CFO’s willingness to support the celebrity CEO’s earnings management preferences. However, this effect is significantly attenuated in firms with high levels of analyst coverage. These findings demonstrate the importance of internal relational alignment and external scrutiny in shaping the influence of high-status CEOs on corporate financial outcomes.

This study contributes to upper echelons theory and earnings management research by highlighting the interplay between the CEO and the CFO in the financial reporting process. While prior research has shown that diversity among top executives affects various firm decisions, relatively little attention has been paid to the CEO-CFO dyad. Moreover, although the individual influence of either the CEO or the CFO on earnings management has been well documented, few studies have examined their joint effect. In this regard, Baker et al. (2019) and Feng et al. (2011) clearly indicate that the relational aspect between CEOs and CFOs is a critical issue that needs to be considered for firms’ financial reporting quality.

From a theoretical perspective, while a substantial body of research grounded in upper echelons theory has emphasized how individual characteristics of top executives shape organizational outcomes, our study extends this perspective by demonstrating that leadership outcomes, including earnings management, are not simply the result of a CEO’s personal characteristics or status, but instead emerge through a socially embedded process involving (1) interpersonal similarity between the CEO and the CFO, and (2) external monitoring by analysts. This relational-contextual framing moves beyond trait-based accounts of leadership and contributes to recent efforts in strategic leadership to understand executive influence as an interactional process shaped by both internal dynamics and external constraints (e.g., Neely Jr et al., 2020).

In addition to these novel contributions, our baseline analysis confirms that a celebrity CEO’s self-interest in preserving their status motivates engagement in earnings management (Li et al., 2022; Malmendier & Tate, 2009). Prior research documents various motives for opportunistic earnings management—such as increasing stock-based pay, extending tenure, and attracting investors (Bergstresser & Philippon, 2006; Burns & Kedia, 2006; Fudenberg & Tirole, 1995; Graham et al., 2005; Welch & Wong, 1998); more recent studies have focused on personality attributes of decision-makers, including demographic characteristics (Ge et al., 2011), overconfidence (Ahmed & Duellman, 2013; Hribar & Yang, 2015), optimism (Bouwman, 2014), and narcissism (Ham et al., 2017). Our findings complement this literature by providing empirical evidence that celebrity CEOs’ desire to protect their status is associated with greater earnings management in their firms.

Our additional analysis further reveals that celebrity CEOs tend to engage in negative earnings management (i.e., income-decreasing discretionary accruals) rather than inflating earnings upward. This pattern is consistent with the incentive to smooth earnings over time by shifting excess profits from periods of strong performance to future periods (Fudenberg & Tirole, 1995). Such income-decreasing earnings management during strong financial performance is well documented; for instance, a survey of 515 auditors from a Big 5 accounting firm found that about 40% of earnings management cases in 1998, a year historically known for one of the biggest bull markets, involved income-decreasing adjustments (Nelson et al., 2002). The prevalence of income-decreasing accruals in our study suggests that celebrity CEOs, concerned about future underperformance and potential reputational damage, may manage earnings to create a buffer for future periods.

Beyond theoretical contributions, this study provides practical implications for corporate governance, particularly for boards of directors who are most responsible for TMT selection and oversight. Since CEOs wield a significant influence over CFO compensation, promotion, and retention decisions, they may exert pressure on CFOs to engage in earnings management to advance their own interests (Dichev et al., 2013; Friedman, 2014). In addition, our findings suggest that the CFO’s role as a guardian of the firm’s financial health may be compromised by their voluntary compliance with the CEO’s decisions, especially when demographic similarity is high between them. Prior research shows that demographic differences between CEOs and directors reduce their interdependence, which can enhance board oversight of the CEO (Zhu & Westphal, 2014). Our findings suggest that increasing demographic differences between CEOs and CFOs may similarly foster greater CFO independence in the financial reporting process.

Therefore, it may be important for the board to actively oversee the selection of the CFO and to prioritize candidates who are demographically dissimilar from the CEO, particularly when the CEO is a recognized celebrity. In addition, our evidence on the mitigating role of analyst coverage suggests that boards should ensure adequate external monitoring by financial analysts to further strengthen CFO independence. These insights underscore the importance of designing governance systems that not only strengthen formal oversight mechanisms but also account for the informal relational dynamics within the TMT and external monitoring pressures. Firms may need to adopt a more holistic governance approach (Aguilera & Ruiz Castillo, 2025)—board oversight, internal executive relationships, external market monitors, and evolving stakeholder expectations—to address the complex, multi-level nature of organizational accountability.

Our analysis is subject to some caveats. First, the theorizing that increasing CEO-CFO demographic similarity causes CFOs to voluntarily comply with requests by CEOs to engage in earnings management may be limited because we did not directly observe the CFOs’ willingness to comply. As is common in archival research, our study infers CFO compliance behavior indirectly without directly measuring whether CFOs actually complied with CEOs’ requests. We acknowledge this as a potential limitation since CFOs may engage in earnings management to seek their own financial gains rather than from voluntary compliance with CEOs’ directives. As suggested by Jiang et al. (2010), CFOs’ equity incentives may be an important consideration in explaining a firm’s earnings management. Therefore, it would be useful if future research could directly measure CFOs’ motivation to engage in earnings management and control for the effect of a broader set of CFOs’ incentives on a firm’s earnings management.

Second, a celebrity CEO’s impression management motives may be contingent on the extent to which a potential threat to celebrity status is perceived. Therefore, it would be worthwhile to further explore additional external factors that may aggravate a celebrity CEO’s perception of potential threats to his or her status and reputation. For example, a high level of firm-specific uncertainty (Gulati et al., 2009) may be a direct threat to a celebrity CEO’s reputation and social status. Celebrity CEOs are under additional pressure to maintain high firm performance, and earnings management may seem like a viable and readily available option for preparing for the future. Higher firm-specific uncertainty makes it more difficult to ensure future performance, increasing the likelihood of earnings management by celebrity CEOs.

Third, while CEO celebrity is plausibly exogenous in the short term—often arising from past achievements or media attention—it may still correlate with unobserved, time-invariant CEO characteristics such as charisma, narcissism, risk tolerance, or strategic orientation that could independently affect financial reporting choices. For example, narcissistic CEOs tend to have an inflated sense of their own capabilities and a strong desire for admiration and praise, which often drives them to engage in impression management behaviors to maintain or enhance their public image (Chatterjee & Hambrick, 2007, 2011; Petrenko et al., 2016). Such narcissistic tendencies can lead CEOs to pursue aggressive strategies that attract media attention, which may also manifest in more opportunistic financial reporting to sustain a favorable public reputation. This potential omitted variable bias may limit the strength of our causal inference. We therefore encourage future research to address this concern by applying alternative identification strategies, such as instrumental variable approaches, to further corroborate the robustness of our findings.

Fourth, although we argued that celebrity CEOs tend to engage in earnings management in order to better manage impressions and maintain their celebrity status, we did not investigate whether celebrity CEOs’ greater engagement in earnings management would actually benefit the CEOs reputation in the long run. Given the potential negative influence of poor earnings quality on firms and the CEO’s reputation, it would be meaningful to explicitly examine, using longitudinal data, whether the use of earnings management increases the likelihood that a CEO maintains celebrity status over the long term.

Finally, our sample spans the 1997–2011 period, which may raise concerns regarding the generalizability of our findings to more recent corporate environments. Although we believe that the core mechanisms examined in our study—such as impression management and CEO-CFO relational dynamics—are grounded in enduring aspects of executive decision-making, it is possible that changes in internal governance practices, external monitoring environments, or business media coverage since 2011 may influence how these mechanisms operate. We therefore encourage future research to revisit our theoretical framework using updated data to evaluate the robustness and contemporary relevance of the observed relationships.

Despite the limitations acknowledged, our study makes a meaningful contribution by showing how the preferences of executives—particularly celebrity CEOs—are enacted within organizations. By emphasizing the relational and contextual contingencies that shape this enactment process—including CEO-CFO demographic similarity and analyst coverage—we offer a more socially embedded perspective on earnings management. This perspective aligns with recent calls in strategic leadership research for greater attention to the relational dynamics that underpin executive influence.

Building on recent discussions of corporate governance, Aguilera and Ruiz Castillo (2025) emphasize the need for governance frameworks that accommodate organizational complexity, evolving stakeholder expectations, and multi-level mechanisms of accountability. Our findings contribute to this broader conversation by showing how internal relational dynamics—specifically CEO-CFO demographic similarity—and external scrutiny through analyst coverage operate as socially embedded governance mechanisms that shape financial reporting behavior. This evidence illustrates the interplay of internal alignment and external monitoring that Aguilera and Ruiz Castillo (2025) highlight as essential features of an updated governance paradigm. At the same time, our study offers a micro-level complement to their vision of the “board of the future.” While they highlight the value of boards that combine insider knowledge, long-tenured experience, and diverse perspectives to navigate uncertainty, our results reveal how insider expertise and interpersonal familiarity can also create channels through which a powerful CEO’s preferences are enacted—unless balanced by robust external oversight.

By documenting this interplay between executive relationships and market monitoring, our work underscores the importance of designing future boards that not only leverage insider insight and long-term perspective, but also maintain sufficient independence and accountability to safeguard the integrity of financial reporting. Future research could extend this relational lens by examining whether and how similar dynamics emerge in interactions among other governance actors and stakeholders—such as boards, institutional investors, customers, employees, or suppliers. Such investigations would advance governance research by clarifying how the interplay of internal executive relationships, external oversight mechanisms, and broader stakeholder interactions generates both constraints and opportunities for organizational decision-making in complex environments.

Beyond this study’s focus on earnings management, the relational-contextual perspective centered on executive dyads and external monitoring could also be applied to other domains—such as risk-taking, innovation, corporate social responsibility, or strategic disclosure—where similar relational and contextual forces may influence executive behavior. We encourage scholars to build on this approach to gain a deeper understanding of how top executives impact firm outcomes not in isolation, but through interactions with key organizational actors and institutional audiences.

Footnotes

ORCID iDs

Gilsoo Lee

Eun Kyung Lee

Ethical considerations

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Notes

References

Abadie

Imbens

G. W.

(2011). Bias-corrected matching estimators for average treatment effects. Journal of Business & Economic Statistics, 29, 1–11.

Aguilera

R. V.

Ruiz Castillo

(2025). Toward an updated corporate governance framework: Fundamentals, disruptions, and future research. BRQ Business Research Quarterly, 28, 336–348.

Ahmed

A. S.

Duellman

(2013). Managerial overconfidence and accounting conservatism. Journal of Accounting Research, 51, 1–30.

Ali

Zhang

(2015). CEO tenure and earnings management. Journal of Accounting and Economics, 59, 60–79.

Arthaud-Day

M. L.

Certo

S. T.

Dalton

C. M.

Dalton

D. R.

(2006). A changing of the guard: Executive and director turnover following corporate financial restatements. Academy of Management Journal, 49, 1119–1136.

Baker

T. A.

Lopez

T. J.

Reitenga

A. L.

Ruch

G. W.

(2019). The influence of CEO and CFO power on accruals and real earnings management. Review of Quantitative Finance and Accounting, 52, 325–345.

Baron

R. A.

(1971). Behavioral effects of interpersonal attraction: Compliance with requests from liked and disliked others. Psychonomic Science, 25, 325–326.

Belliveau

M. A.

O’Reilly

C. A.

III. Wade

J. B.

(1996). Social capital at the top: Effects of social similarity and status on CEO compensation. Academy of Management Journal, 39, 1568–1593.

Bergstresser

Philippon

(2006). CEO incentives and earnings management. Journal of Financial Economics, 80, 511–529.

10.

Bolino

M. C.

Kacmar

K. M.

Turnley

W. H.

Gilstrap

J. B.

(2008). A multi-level review of impression management motives and behaviors. Journal of Management, 34, 1080–1109.

11.

Bouwman

C. H. S.

(2014). Managerial optimism and earnings smoothing. Journal of Banking & Finance, 41, 283–303.

12.

Bozeman

D. P.

Kacmar

K. M.

(1997). A cybernetic model of impression management processes in organization. Organizational Behavior and Human Decision Processes, 69, 9–30.

13.

Brauer

Wiersema

(2018). Analyzing analyst research: A review of past coverage and recommendations for future research. Journal of Management, 44, 218–248.

14.

Brickson

(2000). The impact of identity orientation on individual and organizational outcomes in demographically diverse settings. Academy of Management Review, 25, 82–101.

15.

Burgstahler

Dichev

(1997). Earnings management to avoid earnings decreases and losses. Journal of Accounting and Economics, 24, 99–126.

16.

Burns

Kedia

(2006). The impact of performance-based compensation on misreporting. Journal of Financial Economics, 79, 35–67.

17.

Bushman

Piotroski

Smith

(2005). Insider trading restrictions and analysts’ incentives to follow firms. Journal of Finance, 60, 35–66.

18.

Byrne

(1971). The attraction paradigm. Academic Press.

19.

Byrne

(1997). An overview (and underview) of research and theory within the attraction paradigm. Journal of Social and Personal Relationships, 14, 417–431.

20.

Carberry

E. J.

King

B. G.

(2012). Defensive practice adoption in the face of organizational stigma: Impression management and the diffusion of stock option expensing. Journal of Management Studies, 49, 1137–1167.

21.

Chatterjee

Hambrick

D. C.

(2007). It’s all about me: Narcissistic chief executive officers and their effects on company strategy and performance. Administrative Science Quarterly, 52(3), 351–386.

22.

Chatterjee

Hambrick

D. C.

(2011). Executive personality, capability cues, and risk taking: How narcissistic CEOs react to their successes and stumbles. Administrative Science Quarterly, 56, 202–237.

23.

Chen

Harford

Lin

(2015). Do analysts matter for governance? Evidence from natural experiments. Journal of Financial Economics, 115, 383–410.

24.

Chng

D. H. M.

Rodgers

M. S.

Shih

Song

X. B.

(2015). Leaders’ impression management during organizational decline: The roles of publicity, image concerns, and incentive compensation. The Leadership Quarterly, 26, 270–285.

25.

Cho

S. Y.

Arthurs

J. D.

Townsend

D. M.

Miller

D. R.

Barden

J. Q.

(2016). Performance deviations and acquisition premiums: The impact of CEO celebrity on managerial risk-taking. Strategic Management Journal, 37, 2677–2694.

26.

Chung

Firth

Kim

J. B.

(2002). Institutional monitoring and opportunistic earnings management. Journal of Corporate Finance, 8, 29–48.

27.

Cohen

D. A.

Dey

Lys

T. Z.

(2008). Real and accrual-based earnings management in the pre-and post-Sarbanes-Oxley periods. The Accounting Review, 83, 757–787.

28.

Davidson

W. N.

Jiraporn

Kim

Y. S.

Nemec

(2004). Earnings management following duality-creating successions: Ethnostatistics, impression management, and agency theory. Academy of Management Journal, 47, 267–275.

29.

Dawson

J. F.

Richter

A. W.

(2006). Probing three-way interactions in moderated multiple regression: Development and application of a slope difference test. Journal of Applied Psychology, 91, 917–926.

30.

Dechow

P. M.

Sloan

R. G.

Sweeney

A. P.

(1995). Detecting earnings management. The Accounting Review, 70, 193–225.

31.

DeFond

M. L.

Jiambalvo

(1994). Debt covenant violation and manipulation of accruals. Journal of Accounting and Economics, 17, 145–176.

32.

DeFond

M. L.

Park

C. W.

(1997). Smoothing income in anticipation of future earnings. Journal of Accounting and Economics, 23, 115–139.

33.

DeFond

M. L.

Subramanyam

K. R.

(1998). Auditor changes and discretionary accruals. Journal of Accounting and Economics, 25, 35–67.

34.

Dichev

I. D.

Graham

J. R.

Harvey

C. R.

Rajgopal

(2013). Earnings quality: Evidence from the field. Journal of Accounting and Economics, 56, 1–33.

35.

Feng

Luo

Shevlin

(2011). Why do CFOs become involved in material accounting manipulations? Journal of Accounting and Economics, 51, 21–36.

36.

Finkelstein

(1992). Power in top management teams: Dimensions, measurement, and validation. Academy of Management Journal, 35, 505–538.

37.

Finkelstein

Hambrick

D. C.

Cannella

A. A. J.

(2009). Strategic leadership: Theory and research on executives, top management teams, and boards. Oxford University Press.

38.

Florackis

Sainani

(2021). Can CFOs resist undue pressure from CEOs to manage earnings? Journal of Corporate Finance, 67, Article 101859.

39.

Flynn

F. J.

(2005). Identity orientations and forms of social exchange in organizations. Academy of Management Review, 30, 737–750.

40.

Friedman

H. L.

(2014). Implications of power: When the CEO can pressure the CFO to bias reports. Journal of Accounting and Economics, 58, 117–141.

41.

Fudenberg

Tirole

(1995). A theory of income and dividend smoothing based on incumbency rents. Journal of Political Economy, 103, 75–93.

42.

Gaver

J. J.

Gaver

K. M.

Austin

J. R.

(1995). Additional evidence on bonus plans and income management. Journal of Accounting and Economics, 19(1), 3–28.

43.

Matsumoto

Zhang

J. L.

(2011). Do CFOs have style? An empirical investigation of the effect of individual CFOs on accounting practices. Contemporary Accounting Research, 28, 1141–1179.

44.

Geiger

M. A.

North

D. S.

(2006). Does hiring a new CFO change things? An investigation of changes in discretionary accruals. The Accounting Review, 81, 781–809.

45.

Gentry

R. J.

Shen

(2013). The impacts of performance relative to analyst forecasts and analyst coverage on firm R&D intensity. Strategic Management Journal, 34, 121–130.

46.

Godfrey

Mather

Ramsay

(2003). Earnings and impression management in financial reports: The case of CEO changes. Abacus, 39, 95–123.

47.

Gounopoulos

Loukopoulos

Zhang

(2024). Do CFO career concerns matter? Evidence from IPO financial reporting outcomes. Journal of Corporate Finance, 87, Article 102626.

48.

Graffin

S. D.

Wade

J. B.

Porac

J. F.

McNamee

R. C.

(2008). The impact of CEO status diffusion on the economic outcomes of other senior managers. Organization Science, 19(3), 457–474.

49.

Graham

J. R.

Harvey

C. R.

Rajgopal

(2005). The economic implications of corporate financial reporting. Journal of Accounting and Economics, 40, 3–73.

50.

Gulati

Lavie

Singh

(2009). The nature of partnering experience and the gains from alliances. Strategic Management Journal, 30, 1213–1233.

51.

Ham

Lang

Seybert

Wang

(2017). CFO narcissism and financial reporting quality. Journal of Accounting Research, 55, 1089–1135.

52.

Harris

Bromiley

(2007). Incentives to cheat: The influence of executive compensation and firm performance on financial misrepresentation. Organization Science, 18, 350–367.

53.

Hayward

M. L.

Rindova

V. P.

Pollock

T. G.

(2004). Believing one’s own press: The causes and consequences of CEO celebrity. Strategic Management Journal, 25, 637–653.

54.

Hazarika

Karpoff

J. M.

Nahata

(2012). Internal corporate governance, CEO turnover, and earnings management. Journal of Financial Economics, 104, 44–69.

55.

Healy

P. M.

Palepu

K. G.

(2001). Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of Accounting and Economics, 31, 405–440.

56.

Hennes

Leone

Miller

(2008). The importance of distinguishing errors from irregularities in restatement research: The case of restatements and CEO/CFO turnover. The Accounting Review, 83, 1487–1519.

57.

Hribar

Yang

(2015). CEO overconfidence and management forecasting. Contemporary Accounting Research, 33, 204–227.

58.

Huang

Iun

(2006). The impact of subordinate–supervisor similarity in growth-need strength on work outcomes: The mediating role of perceived similarity. Journal of Organizational Behavior, 27, 1121–1148.

59.

Jensen

M. C.

Meckling

W. H.

(1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305–360.

60.

Jiang

Petroni

Wang

(2010). CFOs and CEOs: Who have the most influence on earnings management? Journal of Financial Economics, 96, 513–526.

61.

Harjoto

M. A.

(2011). Corporate governance and firm value: The impact of corporate social responsibility. Journal of Business Ethics, 103, 351–383.

62.

Johnson

H. H.

Umphress

E. E.

(2019). To help my supervisor: Identification, moral identity, and unethical pro-supervisor behavior. Journal of Business Ethics, 159(2), 519–534.

63.

Jones

J. J.

(1991). Earnings management during import relief investigations. Journal of Accounting Research, 29, 193–228.

64.

Koh

(2011). Value or Glamour? An empirical investigation of the effect of celebrity CEOs on financial reporting practices and firm performance. Accounting & Finance, 51, 517–547.

65.

Kothari

S. P.

Leone

A. J.

Wasley

C. E.

(2005). Performance matched discretionary accrual measures. Journal of Accounting and Economics, 39, 163–197.

66.

Lang

Raedy

J. S.

Wilson

(2006). Earnings management and cross listing: Are reconciled earnings comparable to US earnings? Journal of Accounting and Economics, 42, 255–283.

67.

Leavitt

Sluss

D. M.

(2015). Lying for who we are: An identity-based model of workplace dishonesty. Academy of Management Review, 40(4), 587–610.

68.

Lee

Cho

Arthurs

Lee

E. K.

(2020). Celebrity CEO, identity threat, and impression management: Impact of celebrity status on corporate social responsibility. Journal of Business Research, 111, 69–84.

69.

Shi

Connelly

B. L.

Qin

(2022). CEO awards and financial misconduct. Journal of Management, 48(2), 380–409.

70.

Liang

K. Y.

Zeger

S. L.

(1986). Longitudinal data analysis using generalized linear models. Biometrika, 73, 13–22.

71.

Loi

Ngo

H. Y.

(2009). Work outcomes of relational demography in Chinese vertical dyads. The International Journal of Human Resource Management, 20, 1704–1719.

72.

Malmendier

Pezone

Zheng

(2023). Managerial duties and managerial biases. Management Science, 69(6), 3174–3201.

73.

Malmendier

Tate

(2005). CEO overconfidence and corporate investment. The Journal of Finance, 60, 2661–2700.

74.

Malmendier

Tate

(2009). Superstar CEOs. The Quarterly Journal of Economics, 124, 1593–1638.

75.

McPherson

M. L.

Smith-Lovin

Cook

J. M.

(2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27, 415–444.

76.

Merkl-Davies

Brennan

(2007). Discretionary disclosure strategies in corporate narratives: Incremental information or impression management? Journal of Accounting Literature, 26, 116–196.

77.

Mian

(2001). On the choice and replacement of chief financial officer. Journal of Financial Economics, 60, 143–175.

78.

Montoya

R. M.

Horton

R. S.

(2004). On the importance of cognitive evaluation as a determinant of interpersonal attraction. Journal of Personality and Social Psychology, 86, 696–712.

79.

Myers

J. N.

Myers

L. A.

Skinner

D. J.

(2007). Earnings momentum and earnings management. Journal of Accounting, Auditing & Finance, 22, 249–284.

80.

Neely Jr

B. H.

Lovelace

J. B.

Cowen

A. P.

Hiller

N. J.

(2020). Metacritiques of upper echelons theory: Verdicts and recommendations for future research. Journal of Management, 46(6), 1029–1062.

81.

Nelson

M. W.

Elliott

J. A.

Tarpley

R. L.

(2002). Evidence from auditors about managers’ and auditors’ earnings management decisions. The Accounting Review, 77, 175–202.

82.

Neter

Kutner

M. H.

Nachtsheim

C. J.

Wasserman

(1996). Applied linear statistical models. Irwin.

83.

Peccei

Lee

H. J.

(2005). The impact of gender similarity on employee satisfaction at work: A review and re-evaluation. Journal of Management Studies, 42, 1571–1592.

84.

Petrenko

O. V.

Aime

Ridge

Hill

(2016). Corporate social responsibility or CEO narcissism? CSR motivations and organizational performance. Strategic Management Journal, 37(2), 262–279.

85.

Pollock

T. G.

Ragozzino

R. P.

Blevins

(2024). Not like the rest of us? How CEO celebrity affects quarterly earnings call language. Journal of Management, 50(4), 1198–1229.

86.

Putra

A. A.

Setiawan

(2024). Do CEO characteristics affect earnings management? Corporate Governance: The International Journal of Business in Society, 24(5), 1137–1155.

87.

Qiao

Adegbite

Nguyen

T. H.

(2023). Chief financial officer overconfidence and earnings management. Accounting Forum, 49(1), 181–205.

88.

Ragins

B. R.

(1997). Antecedents of diversified mentoring relationships. Journal of Vocational Behavior, 51, 90–109.

89.

Schlenker

B. R.

(1980). Impression management: The self-concept, social identity, and interpersonal relations. Brooks/Cole.

90.

Schulz

A. C.

Wiersema

M. F.

(2018). The impact of earnings expectations on corporate downsizing. Strategic Management Journal, 39, 2691–2702.

91.

Shao

Zhao

Wang

(2022). Celebrity CEOs and firm innovation investment: Evidence from Chinese-listed companies. Frontiers in Psychology, 13, Article 978946.

92.

Shi

Zhang

Hoskisson

R. E.

(2017). Ripple effects of CEO awards: Investigating the acquisition activities of superstar CEOs’ competitors. Strategic Management Journal, 38(10), 2080–2102.

93.

Shore

L. M.

Cleveland

J. N.

Goldberg

C. B.

(2003). Work attitudes and decisions as a function of manager age and employee age. Journal of Applied Psychology, 88, 529–537.

94.

Skinner

D. J.

Sloan

R. G.

(2002). Earnings surprises, growth expectations, and stock returns or don’t let an earnings torpedo sink your portfolio. Review of Accounting Studies, 7, 289–312.

95.

Sluss

D. M.

Ashforth

B. E.

(2007). Relational identity and identification: Defining ourselves through work relationships. Academy of Management Review, 32(1), 9–32.

96.

Stangor

Lynch

Duan

Glass

(1992). Categorization of individuals on the basis of multiple social features. Journal of Personality and Social Psychology, 62, 207–218.

97.

Steffens

N. K.

Haslam

S. A.

Reicher

S. D.

(2014). Up close and personal: Evidence that shared social identity is a basis for the ‘special’ relationship that binds followers to leaders. The Leadership Quarterly, 25(2), 296–313.

98.

Tajfel

(1982). Social psychology of inter-group relations. Annual Review of Psychology, 33, 1–39.

99.

Tajfel

Turner

J. C.

(1986). The social identity theory of intergroup behavior. In Worchel

Austin

W. G.

(Eds.), Psychology of intergroup relations (pp. 276–293). Nelson–Hall.

100.

Tsui

A. S.

O’reilly

I. I. I. C. A

. (1989). Beyond simple demographic effects: The importance of relational demography in superior-subordinate dyads. Academy of Management Journal, 32, 402–423.

101.

Tsui

A. S.

Porter

L. W.

Egan

T. D.

(2002). When both similarities and dissimilarities matter: Extending the concept of relational demography. Human Relations, 55, 899–929.

102.

Tsui

A. S.

Xin

K. R.

Egan

T. D.

(1995). Relational demography: The missing link in vertical dyad linkage. In Jackson

S. E.

Ruderman

M. N.

(Eds.), Diversity in work teams: Research paradigms for a changing workplace (pp. 97–129). American Psychological Association.

103.

Wade

J. B.

Porac

J. F.

Pollock

T. G.

Graffin

S. D.

(2006). The burden of celebrity: The impact of CEO certification contests on CEO pay and performance. Academy of Management Journal, 49, 643–660.

104.

Watts

Zimmerman

(1986). Positive theory of accounting. Prentice-Hall.

105.

Welch

Wong

(1998). Public administration in a global context: Bridging the gaps of theory and practice between western and non-western wations. Public Administration Review, 58, 40–49.

106.

Westphal

J. D.

Zajac

E. J.

(1995). Who shall govern? CEO/board power, demographic similarity, and new director selection. Administrative Science Quarterly, 40, 60–83.

107.

White

(1980). A heteroskedasticity-consistent covariance matrix and a direct test for heteroskedasticity. Econometrica: Journal of the Econometric Society, 48, 817–838.

108.

Wiersema

M. F.

Zhang

(2011). CEO dismissal: The role of investment analysts. Strategic Management Journal, 32(11), 1161–1182.

109.

Young

M. N.

Buchholtz

A. K.

(2002). Firm performance and CEO pay: Relational demography as a moderator. Journal of Managerial Issues, 14, 296–313.

110.

F. F.

(2008). Analyst coverage and earnings management. Journal of Financial Economics, 88, 245–271.

111.

Zhang

Bartol

K. M.

Smith

K. G.

Pfarrer

M. D.

Khanin

D. M.

(2008). CEOs on the edge: Earnings manipulation and stock-based incentive misalignment. Academy of Management Journal, 51, 241–258.

112.

Zhong

Sluss

D. M.

Badura

K. L.

(2024). Subordinate-to-supervisor relational identification: A meta-analytic review. Journal of Applied Psychology.

113.

Zhou

Long

Yao

(2023). Celebrity CEOs and corporate investment: A psychological contract perspective. International Review of Financial Analysis, 87, Article 102636.

114.

Zhu

D. H.

Westphal

J. D.

(2014). How directors’ prior experience with other demographically similar CEOs affects their appointments onto corporate boards and the consequences for CEO compensation. Academy of Management Journal, 57, 791–813.